Chris Testa-O’Neill

chris@coeo.com

@ctesta_oneill

Introducing Analysis Services

Getting the data

Working with Dimensions Hierarchies

Attribute Relationships

Creating the Cube Working with Measure and Measures Group

Partitions and aggregation design

Browsing the cube

Full exploration of cube properties

Use of additional SSAS components

Calculations (MDX)

Key Performance Indicators

Translations

Perspectives

Administration and Maintenance

Security

Components of SQL Server used for querying and analysing data

Multi-Dimensional is very much alive. Tabular provides new opportunities

Typically uses a data warehouse as it source data Dimension tables

Fact tables

Core object is a cube storing detailed and pre aggregated data

Number of clients can be used to retrieve cube data

Relational Reporting (OLTP) Use of Normalised tables to query data

Can be slow as number of tables used increases or a requirement for aggregate data – Tabular addresses this.

Online Analytical Processing (OLAP) Database type that stores one or more cubes that stores

data in a central repository for reporting purposes

Data Mining Uses OLAP database to explore trands and patterns in the

data

Data Sources provides the connection information. Server Name

Authentication

Database

Data Source Views allows you to define a subset of data from the data source Data Source Wizard

Data Source Designer

Provide contextual information for data in a cube Typically maps to the data in a dimension table of a data warehouse Dimensions form the cube axis Can selectively add attributes to meet business requirements Key properties include Key Columns Name Colums Order by

Time Dimensions

Improves the readability of large dimension data

Adds levels to dimension data so users can drill down into the data

Types of Hierarchies include

Balanced (Natural) Hierarchies

Parent Child

Ragged Hierarchies

Defines relationships that exists between attributes in a dimension

By default, all attributes have a relationship to the key attribute in a star schema

Modifying the default behaviour can Result in more effective aggregation designs

Increases query performance

Reduce memory requirements for processing dimensions

Use Attribute relationships tab in SQL Server 2008

Use the Dimension tab in SQL Server 2005

Cube wizard

Existing Tables

Existing Dimension

Empty Cube

Wizard Capabilities differ from SQL Server 2005 and 2008/2012

Measures are the business metrics stored within the cube

Typically map to measures in a Fact table in a data warehouse

Can create derived measure using MDX expressions

Aggregate property in Measures has additivity issues

Storage Mode property: MOLAP, ROLAP and HOLAP

Measures Group typically map to fact tables

Measures Groups group measures together

Measures Group maps the measure to dimensions

Enterprise Edition

Spread the data across multiple physical disks Improved query performance

Reduced cube processing time

Determine the storage mode on a per partition basis

Design aggregation Enables you to set aggregations based on disk and

performance limit

Usage Based Optimisation a better method

Cube Browser in BIDS

Microsoft Excel

SQL Server Reporting Services

PerformancePoint\Sharepoint

Chris Testa-O’Neill

Chris@Coeo.com

@ctesta_oneill